Comparator

Antelope DPU

for nanosatellites

Leopard DPU

for mirco- and minisatellites

Lion DPU

for small satellites

TRL

9 (from Nov 2023)

Form-factor

CubeSats PC-104

SpaceVPX 3U (mass up to 2.5 kg)

Size

13.24 x 95.9 x 90.1 mm

Version with single Processing Nodes: 90.17 x 95.86 x 50.00 mm
Version with dual Processing Nodes: 90.17 x 95.86 x 78.00 mm

160 mm x 100 mm x 50 mm

Mission size

2U+ or up to 10 kg

3U+

3U+ or up to 500 kg

ESA class mission

IV and V

Mass

Around 130-150 g

Version with single Processing Nodes: <900 g
Version with dual Processing Nodes: <1200 g

Around 2.5 kg

Processing Cores*

Zynq UltraScale+ MPSoC ZU3CG or ZU4CG or ZU5CG:

Dual ARM Cortex-A53 CPU up to 1.5 GHz
Dual ARM Cortex-R5 in lock-step
FPGA for custom function implementation

Zynq UltraScale+ ZU9EG or ZU15EG:

Quad ARM Cortex-A53 CPU up to 1.5 GHz
Dual ARM Cortex-R5 in lock-step
FPGA for custom function implementation

Xilinx Kintex Ultrascale FPGA KU040

Supervisor

RM57 Herkules microcontroller:
Dual 300 MHz ARM Cortex-R5F with FPU in lock-step

ProASIC FPGA and VORAGO microcontroller

SmartFusion2 SoC FPGA, ARM Cortex-M3, 86KLEs

Redundancy

Not available

Available redundancy options

Not available

Memory DPU*

8 GB DDR4 with ECC

4-16 GiB DDR4 with ECC
4-16 GiB SLC flash-based file system storage (EDAC)
Up to 2x256 GiB SLC flash-based data storage

2-8 GiB DDR4 with ECC
4-32 GiB SLC flash-based file system storage (EDAC)
Up to 4x256 GiB of SLC flash-based data storage

Memory supervisor

4 GB SLC NAND Flash

Not available

Program Memory Size: 512 kB
Data RAM Size: 64 kB

Available interfaces

CAN, I2C, GPIO, SPI, RS422/485, UART, GPS PPS, LVDS or GTH transceivers, USB 3.0, USB 2.0, Ethernet 1Gb, SATA

CAN, LVDS, SPI, RS422/485, UART, GTY and GTH transceivers

CAN, LVDS, RS422/485, UART, GTH transceiversLVDS/RS422 interfaces compatible with X/S-Band radios

Supply voltage

From 6 to 14 V (VBAT)
5V regulated

From 6.5 to 14 V (VBAT)

12V to use integrated 12V to 5V converter
5V for direct power supply

Power consumption

From 5 W to 10W – depending on workload and specified processing speed

From 7.5 W to 40 W – depending on workload and specified processing speed

Up to 20W - depending on workload and specified processing speed

Computational Throughput for Neural Networks*

160 GOPS

3 TOPS

1516 GOPS

Operating Temperature

-40 to 85°C

Software ecosystem*

64-bit Linux
bare-metal applications

64-bit Linux
Deep Learning Accelerator fed with PyTorch, Caffe or TensorFlow models

FreeRTOS (default) or RTEMS

Book now

Antelope DPU

for nanosatellites

Leopard DPU

for mirco- and minisatellites

TRL

9 (from Nov 2023)

Form-factor

CubeSats PC-104

Size

13.24 x 95.9 x 90.1 mm

Version with single Processing Nodes: 90.17 x 95.86 x 50.00 mm
Version with dual Processing Nodes: 90.17 x 95.86 x 78.00 mm

Mission size

2U+ or up to 10 kg

3U+

ESA class mission

IV and V

Mass

Around 130-150 g

Version with single Processing Nodes: <900 g
Version with dual Processing Nodes: <1200 g

Processing Cores*

Zynq UltraScale+ MPSoC ZU3CG or ZU4CG or ZU5CG:

Dual ARM Cortex-A53 CPU up to 1.5 GHz
Dual ARM Cortex-R5 in lock-step
FPGA for custom function implementation

Zynq UltraScale+ ZU9EG or ZU15EG:

Quad ARM Cortex-A53 CPU up to 1.5 GHz
Dual ARM Cortex-R5 in lock-step
FPGA for custom function implementation

Supervisor

RM57 Herkules microcontroller:
Dual 300 MHz ARM Cortex-R5F with FPU in lock-step

ProASIC FPGA and VORAGO microcontroller

Redundancy

Not available

Available redundancy options

Memory DPU*

8 GB DDR4 with ECC

4-16 GiB DDR4 with ECC
4-16 GiB SLC flash-based file system storage (EDAC)
Up to 2x256 GiB SLC flash-based data storage

Memory supervisor

4 GB SLC NAND Flash

Not available

Available interfaces

CAN, I2C, GPIO, SPI, RS422/485, UART, GPS PPS, LVDS or GTH transceivers, USB 3.0, USB 2.0, Ethernet 1Gb, SATA

CAN, LVDS, SPI, RS422/485, UART, GTY and GTH transceivers

Supply voltage

From 6 to 14 V (VBAT)
5V regulated

From 6.5 to 14 V (VBAT)

Power consumption

From 5 W to 10W – depending on workload and specified processing speed

From 7.5 W to 40 W – depending on workload and specified processing speed

Computational Throughput for Neural Networks*

160 GOPS

3 TOPS

Operating Temperature

-40 to 85°C

Software ecosystem*

64-bit Linux
bare-metal applications

64-bit Linux
Deep Learning Accelerator fed with PyTorch, Caffe or TensorFlow models

Book now

Antelope DPU

for nanosatellites

Lion DPU

for small satellites

TRL

9 (from Nov 2023)

Form-factor

CubeSats PC-104

SpaceVPX 3U (mass up to 2.5 kg)

Size

13.24 x 95.9 x 90.1 mm

160 mm x 100 mm x 50 mm

Mission size

2U+ or up to 10 kg

3U+ or up to 500 kg

ESA class mission

Mass

Around 130-150 g

Around 2.5 kg

Processing Cores*

Zynq UltraScale+ MPSoC ZU3CG or ZU4CG or ZU5CG:

Dual ARM Cortex-A53 CPU up to 1.5 GHz
Dual ARM Cortex-R5 in lock-step
FPGA for custom function implementation

Xilinx Kintex Ultrascale FPGA KU040

Supervisor

RM57 Herkules microcontroller:
Dual 300 MHz ARM Cortex-R5F with FPU in lock-step

SmartFusion2 SoC FPGA, ARM Cortex-M3, 86KLEs

Redundancy

Not available

Memory DPU*

8 GB DDR4 with ECC

2-8 GiB DDR4 with ECC
4-32 GiB SLC flash-based file system storage (EDAC)
Up to 4x256 GiB of SLC flash-based data storage

Memory supervisor

4 GB SLC NAND Flash

Program Memory Size: 512 kB
Data RAM Size: 64 kB

Available interfaces

CAN, I2C, GPIO, SPI, RS422/485, UART, GPS PPS, LVDS or GTH transceivers, USB 3.0, USB 2.0, Ethernet 1Gb, SATA

CAN, LVDS, RS422/485, UART, GTH transceiversLVDS/RS422 interfaces compatible with X/S-Band radios

Supply voltage

From 6 to 14 V (VBAT)
5V regulated

12V to use integrated 12V to 5V converter
5V for direct power supply

Power consumption

From 5 W to 10W – depending on workload and specified processing speed

Up to 20W - depending on workload and specified processing speed

Computational Throughput for Neural Networks*

160 GOPS

1516 GOPS

Operating Temperature

-40 to 85°C

Software ecosystem*

64-bit Linux
bare-metal applications

FreeRTOS (default) or RTEMS

Book now

Leopard DPU

for mirco- and minisatellites

Lion DPU

for small satellites

TRL

9 (from Nov 2023)

Form-factor

CubeSats PC-104

SpaceVPX 3U (mass up to 2.5 kg)

Size

Version with single Processing Nodes: 90.17 x 95.86 x 50.00 mm
Version with dual Processing Nodes: 90.17 x 95.86 x 78.00 mm

160 mm x 100 mm x 50 mm

Mission size

3U+

3U+ or up to 500 kg

ESA class mission

IV and V

Mass

Version with single Processing Nodes: <900 g
Version with dual Processing Nodes: <1200 g

Around 2.5 kg

Processing Cores*

Zynq UltraScale+ ZU9EG or ZU15EG:

Quad ARM Cortex-A53 CPU up to 1.5 GHz
Dual ARM Cortex-R5 in lock-step
FPGA for custom function implementation

Xilinx Kintex Ultrascale FPGA KU040

Supervisor

ProASIC FPGA and VORAGO microcontroller

SmartFusion2 SoC FPGA, ARM Cortex-M3, 86KLEs

Redundancy

Available redundancy options

Not available

Memory DPU*

4-16 GiB DDR4 with ECC
4-16 GiB SLC flash-based file system storage (EDAC)
Up to 2x256 GiB SLC flash-based data storage

2-8 GiB DDR4 with ECC
4-32 GiB SLC flash-based file system storage (EDAC)
Up to 4x256 GiB of SLC flash-based data storage

Memory supervisor

Not available

Program Memory Size: 512 kB
Data RAM Size: 64 kB

Available interfaces

CAN, LVDS, SPI, RS422/485, UART, GTY and GTH transceivers

CAN, LVDS, RS422/485, UART, GTH transceiversLVDS/RS422 interfaces compatible with X/S-Band radios

Supply voltage

From 6.5 to 14 V (VBAT)

12V to use integrated 12V to 5V converter
5V for direct power supply

Power consumption

From 7.5 W to 40 W – depending on workload and specified processing speed

Up to 20W - depending on workload and specified processing speed

Computational Throughput for Neural Networks*

3 TOPS

1516 GOPS

Operating Temperature

-40 to 85°C

Software ecosystem*

64-bit Linux
Deep Learning Accelerator fed with PyTorch, Caffe or TensorFlow models

FreeRTOS (default) or RTEMS

Book now

Highlighted offerings are inclusive of the Smart Mission Lab services, accessible within a subscription package.

Comparator

Still not sure which one to choose?